Online learning is a subset of distance learning, sufficiently technologically different from traditional instructional television to merit separate discussion. The basic premise of online learning is to electronically connect physically separate teachers, students and learning experiences. Original content and learning experiences are placed online, links can be to anything and anyone on the Internet, communication and other software connects people with each other. By definition, class always meets in a technological, information-rich environment.

Cisco Academyâs e-learning online educational ecosystem is currently implemented for 2500 academies internationally [9]. According to Jim LeValley, Program Manager of Cisco Press, technology enhanced learning was isolated in the 1980s, supplemental in the 1990s and will be integrated in 2000. Cisco e-learning tools include constant feedback, large forum collaboration tools, as well as assessment tools. More than 50 companies sell integrated online learning systems or individual tools to educational institutions. Some universities now offer online degrees. Many offer online courses. Motivations for universities offering online courses (and for students wanting to take online courses) are diverse.

Table 3. Motivations for Online Courses

Some pedagogies are easy to implement online. Not surprisingly, online learning is conducive to multiple- choice, automatically graded exercises and exams. CBT (computer based training) with branching and programmed feedback translates easily to online courses.

Perhaps less intuitively obvious, online courses also excel at seminar discussion formats. Asynchronous communication tools let students in different time zones on different schedules participate more fully in graduate seminar discussions than might occur in a traditional 2 hour classroom seminar where only one person gets to talk at a time. Online professors can divide their class into small discussion groups of 4 or 5 students and guide each group in discussion assignments.

One-to-one professor-student interactions can be initiated by the professor far more easily than in traditional classroom instruction. In the classroom, 20 to 200 other students attend class together. Asking a student a question during class only happens while all of the others watch and wait. A few students stay after class to ask questions. And a few attend office hours. But the instructor is not easily able to ask each individual student how she is doing and what material she would like to see covered in more depth. Online itâs easy to keep in touch with and to keep track of each studentâs progress as an individual. Instant messaging even allows a professor or fellow students to ask and answer questions instantly, providing both parties are online.

Professors and students are both more accountable in an online course. In the classroom where, like a conversation, no record of the teaching persists other than student notes after the hour is up. But in an online course most or all of the content and the discussion exists online and can be viewed and reviewed later. Some courses let students know they will be monitored and then track how many lectures they have listened to. Privacy and ethical considerations guide access to such content. But the potential for review, even if just for self- improvement, is much more detailed than for in-person courses.

Creating course materials is a lot more work in online courses. Perfection matters more and takes longer. Classroom handouts don't need to be designed with elegant navigational consistency. "Getting it just right" is an involved process including both substance and myriad details, which, if done poorly, hurt the overall impact of the course materials. Televised courses faced the same comparison: broadcast television and live university instructional television courses were not at all comparable in quality, production values and polish. Online web pages are compared with big budget corporate web sites, but more often than instructional television, online courses try to rise to the challenge of presenting commercial quality polished products.

Once a course has been created online, it can be refined and improved in subsequent semesters. The first time through tends to be a struggle between creation and interaction ö both aspects take large amounts of time.

Many forms of pedagogies being used in different courses, many of which could be enhanced by Internet 2 technologies and services:

 Table 4. Some Online Pedagogies

While the validity of classroom teaching is rarely questioned, teaching with technology invariably attracts the cautious question: does online learning work? Chickering and Ehrmann distilled education research and collective experience down to seven principles, offering seminars in exactly how to implement them for online learning. Their approach, described below, is more prescriptive than evaluative.

Table 5. How to Teach Online, from Chickering and Ehrmann

• Encourage Contacts Between Students and Faculty
• Develop Reciprocity and Cooperation Among Students
• Use Active Learning Techniques
• Give Prompt Feedback
• Emphasize Time on Task
• Communicate High Expectations
• Respect Diverse Talents and Ways of Learning

Hara and Kling [5] raise the issue of student frustration in online courses, claiming negative aspects of online learning are a taboo topic among researchers and practitioners. Kling suggests the major body of literature on distance education is favorable, and the tendency, once a technological utopian consensus has been reached, is to resist dissonant ideas.

Their case study was a detailed examination of an online course taught by a doctoral student who took over an already developed online class at the last minute because the instructor who developed it became ill. The course used an email listserve rather than online conferencing. The instructor failed to send out assignments on schedule. And the university appeared to lack technical support mechanisms to help online students when they had problems. In other words, it was a "worst case scenario" situation. Lots of student frustration was documented in the six-person online course, and the authors concluded that this frustration inhibited their educational opportunity.

One explanation for why frustration gets ignored by researchers (and in student evaluations) is that frustrating failures feel like a temporary technology problem outside of the actual course. For example, when the network dies in the middle of one studentâs final exam, the exam does not get blamed ö the network is at fault. The locus of fault for frustration with online courses is widely distributed. And the expectation for "things getting better next time" has basis in fact (engineers continually try to fix problems and often succeed) and past experience with technology (this yearâs computer is more powerful than last yearâs computer).

Klingâs suggestion to pay more attention to failures of educational innovation remains well taken. Anecdotes collected at informal phases need to include the bad as well as the good.

Many speakers at the 1999 September Internet2 Sociotechnical Summit called for hard data documenting the effects of I2 (and other) technology. Leiferâs learning lab exemplifies this spirit of constant experimentation [8]. Approximately 50% of course-development resources are spent incorporating elements of learning theory, pedagogy, domain content, and design theory to work with researchers from the content discipline for an online learning course.

The Learning Labâs online courses tend to be international, engaging academic, corporate and cultural institutions around the world to collaborate with residential learners in fields as diverse as mechanical engineering and English literature. Their research demonstrates how small changes in protocol and technology can have profound effects on learning. An example is online multiple-choice exams. When the course developers added a form requiring students to explain why they picked the answer they thought was right, the multiple choice exam transformed into a sophisticated tracking tool to help the instructor know what students understood at any point in the semester.

Leiferâs Learning Lab research also notes how small changes can alter student and faculty time-allocation to the course (sometimes taking more time than they can give). This brings up the often -repeated remonstration that students and teachers are also people, with complex lives, and that taking an online course occurs within that personal context. For both personal context and other individual difference reasons, the effects of technology are not uniform. And, as mentioned previously, every change begets another change, leaving us to study a moving target.